Abstract
Modal parameter identification can be a valuable tool in mechanical engineering to predict vibrational behaviour and avoid machine damage during operation. Operational modal analysis is an output-only identification tool motivated by the structural identification of civil engineering structures, which are excited by ambient conditions. This technique is increasingly applied in mechanical engineering in order to characterise the system behaviour during operation as modal parameters can vary under operating conditions. The following study investigates the application of operational modal analysis on an axial compressor under operating conditions. Since the modal parameters of the system change depending on the life history and during the operation of the system, a corresponding data analysis might allow us to identify the present status of the system. Eigenfrequencies and eigenvectors are studied for the use of structural health monitoring approaches. According to the analysis, eigenfrequencies represent robust parameters for the studied purpose. Eigenvectors are sensitive to damages but need further investigation, especially for rotating machinery. This study will help the user to set up a virtual model, which describes the system behaviour for different boundary conditions. This in turn, will provide an accurate prediction of the vibrational behaviour in order to assure a safe operation.
Highlights
The operation of rotating machinery such as turbomachinery requires nonintrusive monitoring due to economic and accessibility reasons
This study showed that the modal parameters depend on rotational speed when comparing experimental modal analysis (EMA) and operational modal analysis (OMA) results, which vary due to the rotational impact on the structural behaviour
A study on an axial compressor test rig was conducted to qualify the application of operational modal analysis for structural health monitoring purposes on rotating machinery, explicitly turbomachinery
Summary
The operation of rotating machinery such as turbomachinery requires nonintrusive monitoring due to economic and accessibility reasons. A set of control parameters is updated in an iterative manner to detect possible changes in the operational behaviour. This in turn, is defined as structural health monitoring (SHM), which should prevent possible damages and failures from happening. SHM tools usually depend on some kind of analytical and most often purely linear model. Neural networks are used for the decision-making process of the systems’ structural condition as they offer a high reliability when trained with a big data set according to Sohn et al [1]. The authors further explain in their literature review that many insights of monitoring processes are gained under laboratory conditions and need to be transferred to real applications
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call